JP6299544B2 - Solid-state imaging device - Google Patents

Description

  The present invention relates to a solid-state imaging device.

  A solid-state image sensor that performs a global shutter operation with a back-illuminated type has been proposed. In relation to this technique, Patent Document 1 discloses a solid-state imaging device in which a first substrate and a second substrate are electrically connected. In Patent Document 1, a photodiode is disposed on a first substrate, and a memory using a capacitor is provided on a second substrate. A global shutter is realized by recording an optical signal on a capacitor provided on the second substrate.

  Here, as a method of expanding the dynamic range in a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor, a method of combining an optical signal obtained by long-time exposure and an optical signal obtained by short-time exposure There is. If this method is to be realized with a back-illuminated solid-state imaging device provided with a capacitor, an optical signal obtained by long-time exposure and an optical signal obtained by short-time exposure are stored separately. Therefore, it is necessary to increase the number of capacitors.

JP 2012-248952 A

  However, since the area of the substrate is limited, if the number of capacitors is increased, the area per capacitor must be reduced. As a result, the capacitance of the capacitor decreases. Furthermore, if the leakage current is the same, the smaller the capacitance of the capacitor, the greater the degradation of image quality. Therefore, the image quality is degraded by simply increasing the number of capacitors.

  In view of the above problems, an object of the present invention is to provide a solid-state imaging device capable of expanding the dynamic range without increasing the number of capacitors.

  Therefore, the present invention provides photoelectric conversion means for performing photoelectric conversion, first optical signal storage means for storing an optical signal obtained by photoelectric conversion by the photoelectric conversion means, and photoelectric conversion by the photoelectric conversion means. The second optical signal storage means for storing the obtained optical signal, the optical signal stored in the first optical signal storage means and the optical signal stored in the second optical signal storage means are combined. A first optical signal accumulating unit that accumulates the first optical signal obtained by performing the photoelectric conversion in the first period by the photoelectric converting unit, and the second optical signal accumulating unit. The means stores the second optical signal obtained by photoelectric conversion by the photoelectric conversion means in a second period shorter than the first period, and the synthesizing means stores the second optical signal of the first optical signal. When the noise is greater than or equal to the noise of the second optical signal, the first light To provide a solid-state imaging device for combining the No. and the second optical signal.

  According to the present invention, the dynamic range can be expanded without increasing the number of capacitors.

1 is an overall view of a solid-state imaging device according to a first embodiment. 1 is a diagram illustrating an overall configuration of a first substrate of a solid-state imaging device according to a first embodiment; FIG. 2 is a diagram illustrating an overall configuration of a second substrate of the solid-state imaging device according to the first embodiment. It is a figure which shows the cross-sectional schematic diagram of the state in which the 1st board | substrate and the 2nd board | substrate were connected. 1 is a diagram illustrating a pixel structure of a solid-state imaging device according to a first embodiment. FIG. 2 is a diagram illustrating details of a circuit configuration of each pixel of the solid-state imaging device according to the first embodiment. 1 is a diagram showing a configuration of a column processing circuit according to a first exemplary embodiment. FIG. 6 is a timing chart of a global shutter operation according to the first embodiment. 3 is a flowchart showing an operation when a global shutter operation is performed by the column processing circuit according to the first exemplary embodiment; It is the figure which compared the noise of the 1st optical signal, and the noise of the 2nd optical signal. 3 is a graph showing the relationship between the light amount and the optical signal in the first embodiment. FIG. 6 is a timing chart of a rolling shutter operation according to the first embodiment. 3 is a flowchart showing an operation when a rolling shutter operation of the column processing circuit according to the first exemplary embodiment is performed.

(Embodiment 1)
Embodiments of the present invention will be described below with reference to the drawings.
FIG. 1 is an overall view of a solid-state imaging device 100 according to the first embodiment. The solid-state imaging device 100 includes a substrate 1 that is a first substrate, a substrate 2 that is a second substrate, and a substrate 30 that is a third substrate. The substrate 1 has a region where pixels having a circuit for converting light into an optical signal are spread. The substrate 2 has a circuit that receives and processes an optical signal from the substrate 1. The substrate 30 is a heat dissipation plate and has a function of releasing heat generated from the back surface of the substrate 2 to the substrates 1 and 2 to the outside. That is, the heat generated in the substrate 1 is released through the substrate 2. The substrate 30 may be provided with temperature control means that actively radiates heat using, for example, a Peltier element. Further, the surface (front surface) of the substrate 1 and the surface of the substrate 2 face each other. The substrate 30 is provided on the back side of the substrate 2.

  For example, the thickness of the silicon layer of the substrate 1 is about 10 μm. For example, the thickness of the silicon layer of the substrate 2 is about 200 to 1000 μm. For example, the thickness of the substrate 30 is about 1000 to 3000 μm. The material of the substrate 30 may be a metal, a carbon-based material such as ceramic, plastic, or diamond having high thermal conductivity, or sapphire. The substrate 30 may be various semiconductor substrates or a circuit board having a multilayer structure. The substrate 30 may be a part of the package instead of an independent substrate.

  Here, in FIG. 1, regarding the size relationship of the substrate areas, the area of the substrate 2 is larger than the area of the substrate 1, and the area of the substrate 30 is larger than the area of the substrate 2. A pad 2a is provided in a portion of the substrate 2 that protrudes beyond the substrate 1, and a signal or the like is exchanged with an external circuit via the pad 2a. The size relationship of the substrate areas is not limited to the above, and may be the same size. Further, for example, a micro bump may be formed therethrough through a through tunnel from the front surface to the back surface of the substrate 2. In this case, the substrate 1 may be configured to be connected to the substrate 30 using the micro bumps, and may be configured to exchange signals, power, or the like with an external circuit via the substrate 30.

  FIG. 2 is a diagram illustrating an overall configuration of the substrate 1 that is the first substrate of the solid-state imaging device 100 according to the first embodiment. The substrate 1 has a pixel covering area 105, peripheral circuits 101, 102, 103, 104, 106 and a pad 107. In the pixel covering area 105, a part of pixel circuits including photodiodes is regularly arranged. The peripheral circuits 101, 102, 103, and 104 are provided with a drive circuit for driving the pixels. With these drive circuits, each pixel provided in the pixel covering area 105 performs a predetermined operation. The peripheral circuit 106 is provided with a control circuit that performs control so as to appropriately perform the driving described above. Specifically, the control circuit controls which pixel is driven at what timing. The peripheral circuit 106 is also provided with a power supply circuit that stably supplies power. The pad 107 connects the substrate 1 and the substrate 2 and exchanges signals between the substrate 1 and the substrate 2. Furthermore, not only signals but also power exchange is performed via the pad 107.

  FIG. 3 is a diagram illustrating an overall configuration of the substrate 2 that is the second substrate of the solid-state imaging device 100 according to the first embodiment. The substrate 2 includes a pixel covering area 205, peripheral circuits 201, 202, 203, 204, 206 and a pad 207. In the pixel covering area 205, other pixel circuits corresponding to the pixel covering area 105 are regularly arranged. The pixel circuit provided in the pixel covering area 205 is configured to receive and hold an optical signal from the pixel circuit provided in the pixel covering area 105. That is, the pixel circuit provided on the substrate 1 and the pixel circuit provided on the substrate 2 form one pixel circuit.

  The peripheral circuits 201, 202, 203, 204, and 206 include a digital / analog conversion circuit, an analog / digital conversion circuit, an image processing circuit that performs image processing, a power supply circuit, a timing adjustment circuit, a clock signal supply circuit, and the like. The analog / digital conversion circuit is a circuit that performs analog / digital conversion on an optical signal or a reference potential for each row or column. The digital / analog conversion circuit includes a digital / analog converter that generates an intermediate potential signal of each circuit and generates a high-performance ramp waveform necessary for analog / digital conversion. The image processing circuit is a circuit for receiving an image signal from each pixel to form an image. For example, a white defect correction function, a gradation setting function, a white balance setting function, a field memory, an image / video compression circuit, and A circuit for serial communication with the outside is included. The pad 207 connects the substrate 1 and the substrate 2 and exchanges signals between the substrate 1 and the substrate 2.

  Here, FIG.2 and FIG.3 is the figure which looked at the board | substrate 1 and the board | substrate 2 from the front side, respectively. The board | substrate 1 and the board | substrate 2 are arrange | positioned so that front side may mutually oppose. In other words, the substrates 1 and 2 are configured such that one of the substrates shown in FIGS. 2 and 3 is horizontally reversed and connected to the other substrate. Accordingly, the pad 107 and the pad 207 are disposed so as to face each other, and the pad 107 and the pad 207 are connected via the micro bump. Note that the pad and the micro bump are also present between the pixel covering area 105 and the pixel covering area 205, and connect the two.

  FIG. 4 is a schematic cross-sectional view showing a state in which the substrate 1 as the first substrate and the substrate 2 as the second substrate are connected to each other. As described above, the front side of the substrate 1 and the front side of the substrate 2 are arranged to face each other. That is, the upper side of the sectional view shown in FIG. 4 is the back side of the substrate 1, and the lower side is the back side of the substrate 2. The substrate 1 is configured such that a wiring region 5, a silicon region 6, an insulating film 21 such as a silicon oxide film, a color filter 22, and a microlens 23 are laminated from the front side. The substrate 2 is configured so that the wiring region 4 and the silicon region 3 are laminated from the front side.

  First, the structure of the substrate 1 will be described. Light enters from the back surface of the substrate 1. The light passes through the micro lens 23. The color filter 22 allows light having a specific wavelength to pass therethrough. Then, the light that has passed through the color filter 22 passes through the insulating film 21 and enters the photoelectric conversion region 18. Here, the photoelectric conversion region 18 is an n-type semiconductor. The first p + type region 20 prevents the generation of non-optical signal carriers generated at the interface between the photoelectric conversion region 18 and the insulating film 14 and the insulating film 21, and configures the photoelectric conversion region 18 as a buried photodiode. It is an area for. The second p + type region 19 is a region for separating from a photoelectric conversion region of an adjacent pixel. The p-well region 17 is a region for configuring an n-MOSFET (Metal-Oxide-Semiconductor Field-Effect Transistor) used for the pixel. The MOSFET in the pixel includes a p-well region 17, an n + -type region 16 that serves as a source / drain, and a gate electrode 15. The optical signal charge is converted into a voltage by the MOSFET, and the optical signal of the voltage passes through the inside of the wiring region 5 constituted by the insulating film 14 and the metal wiring 13, reaches the connection portion 12, and is transmitted to the substrate 2. The The optical signal of the voltage transmitted to the substrate 2 is transmitted to the MOSFET in the substrate 2 through the inside of the wiring region 4 constituted by the insulating film 11 and the metal wiring 10. The MOSFET of the substrate 2 includes a p-type substrate 7 including a p-well region, a source / drain region 8 and a gate electrode 9. A MOS capacitor is formed using a gate electrode or the like. In this way, the second pixel circuit in the substrate 2 is configured.

  In FIG. 4, only the n-MOSFET is included in the pixels in the substrate 1 and the substrate 2, but only the p-MOSFET may be included. In this case, the conductivity types of the photoelectric conversion region 18, the p well region 17, the n + type region 16, the p type substrate 7 and the source / drain region 8 are all reversed. Further, a p-well region and an n-well region may be formed in the pixel to form a C-MOS circuit. Although only the pixel portion is shown in FIG. 4, CMOS circuits may be used for the peripheral circuits 101 to 104 and 106 provided around the pixel covering area 105 as shown in FIG. 2.

  FIG. 5 is a diagram illustrating a pixel structure of the solid-state imaging device 100 according to the first embodiment. A plurality of pixel portions 50 are provided in the pixel spread area 105 and the pixel spread area 205. The plurality of pixel units 50 are arranged in a matrix as shown in FIG. FIG. 5 is a diagram schematically showing how the pixel units 50 are arranged in a matrix. Actually, as described above with reference to FIG. 4, each pixel unit 50 is physically arranged. It does not have to be arranged separately.

  Each pixel unit 50 is connected to a vertical scanning circuit 77 via a row signal line 70 and connected to a column processing circuit 60 via a column signal line 170. The column processing circuit 60 is connected to the horizontal scanning circuit 177. The vertical scanning circuit 77 is provided in the peripheral circuits 101 and 102 and the peripheral circuits 201 and 202. The column processing circuit 60 and the horizontal scanning circuit 177 are provided in the peripheral circuits 103 and 104 and the peripheral circuits 203 and 204.

  The vertical scanning circuit 77 is composed of, for example, a shift register, and controls the driving of the pixel unit 50 in units of rows. This drive control includes a reset operation, an accumulation operation, a signal readout operation, and the like of the pixel unit 50. In order to perform this drive control, the vertical scanning circuit 77 outputs a control signal (control pulse) to each pixel unit 50 via the row signal line 70 provided for each row, and makes the pixel unit 50 independent for each row. And control. When the vertical scanning circuit 77 performs drive control, a pixel signal is output from the pixel unit 50 to the column signal line 170 provided for each column.

  The column processing circuit 60 processes a signal output from the pixel unit 50 to the column signal line 170. The column processing circuit 60 functions as a combining unit that combines a plurality of optical signals obtained by exposure at a plurality of different timings. Details will be described later. The horizontal scanning circuit 177 is configured by, for example, a shift register, selects a pixel column from which pixel signals are read, sequentially selects the column processing circuit 60 related to the selected pixel column, and sequentially receives signals from the column processing circuit 60. The pixel signal is read out by outputting to the next circuit.

  FIG. 6 is a diagram illustrating details of a circuit configuration of each pixel of the solid-state imaging device 100 according to the first embodiment. In FIG. 6, two pixel portions 50-1 and 50-2 connected to the same column signal line 170 are shown. The connection parts 136 and 156 correspond to the connection part 12 in FIG. 4 and connect the substrate 1 and the substrate 2. A component on the left side of the connection portions 136 and 156 is formed on the substrate 1, and a component on the right side of the connection portions 136 and 156 is formed on the substrate 2. On the substrate 1 side, photoelectric conversion circuits 121 and 122 for converting light into electric signals are formed. On the substrate 2 side, a signal holding circuit that mainly holds an electric signal until an appropriate timing, and an output circuit that outputs the held signal to the column signal line 170 are formed. Details will be described later.

  First, the photoelectric conversion circuits 121 and 122 will be described. The photoelectric conversion circuits 121 and 122 have photodiodes 130 and 150, respectively. The photodiodes 130 and 150 function as photoelectric conversion means for performing photoelectric conversion, and convert incident light into electric charges, that is, hole / electron pairs. Either one of the hole / electron pairs is accumulated in the photodiodes 130 and 150. In the present embodiment, electrons are accumulated. Transfer transistors 131 and 151 are connected between the photodiodes 130 and 150 and the floating diffusions 133 and 153, respectively. The transfer transistors 131 and 151 are controlled by signals φtx1 and φtx2 output from the vertical scanning circuit 77, respectively. When the transfer transistors 131 and 151 are turned on by the signals φtx1 and φtx2, the electrons accumulated in the photodiodes 130 and 150 are transferred to the floating diffusions 133 and 153, respectively.

  The floating diffusions 133 and 153 have a function as second optical signal storage means. The floating diffusions 133 and 153 are a minute capacitance and a floating capacitance formed of a diffusion layer. The potential is reduced by the electrons transferred to the floating diffusions 133 and 153, and the electric charges are converted into electric signals in the floating diffusions 133 and 153.

  Reset transistors 132 and 152 are connected between the floating diffusions 133 and 153 and the power supply voltage Vdd, respectively. The reset transistors 132 and 152 are transistors that reset the floating diffusions 133 and 153, respectively, and are controlled by signals φrst1 and φrst2 output from the vertical scanning circuit 77, respectively. In addition, amplifier transistors 134 and 154 are connected to the floating diffusions 133 and 153, respectively. The amplifier transistors 134 and 154 amplify the potentials of the floating diffusions 133 and 153, respectively. Load transistors 135 and 155 functioning as current sources are connected to the ground sides of the amplifier transistors 134 and 154, respectively. The load transistors 135 and 155 function as loads by applying a predetermined voltage to the gate electrodes of the load transistors 135 and 155, respectively. This predetermined voltage may be, for example, a voltage of about 1.2 V, which is larger than a voltage threshold described later. This voltage value may be variable, thereby changing the current value. The voltage generated in the load is transmitted to the substrate 2 side through the connection portions 136 and 156.

  In the circuit on the substrate 2 side, switch transistors 140 and 160 functioning as switches for turning on and off the electrical connection with the substrate 1 are provided between the connection portions 136 and 156 and the nodes 145 and 165. The switch transistors 140 and 160 are controlled by signals φtx1a and φtx2a output from the vertical scanning circuit 77, respectively. Capacitors 142 and 144 are connected to the switch transistor 140 via switch transistors 141 and 143, respectively. In other words, the switch transistors 141 and 143 are connected between the node 145 and the capacitors 142 and 144, respectively. Capacitors 162 and 164 are connected to the switch transistor 160 via switch transistors 161 and 163, respectively. In other words, the switch transistors 161 and 163 are connected between the node 165 and the capacitors 162 and 164, respectively. The switch transistors 141, 143, 161, and 163 are controlled by signals φtx1b, φtx1c, φtx2b, and φtx2c output from the vertical scanning circuit 77, respectively. The capacitors 142 and 144 and the capacitors 162 and 164 each have a function as first optical signal storage means.

  Reset transistors 146 and 166 are connected between nodes 145 and 165 and power supply voltage Vdd, respectively. The reset transistors 146 and 166 are transistors that reset the nodes 145 and 165, respectively, and are controlled by signals φrst1a and φrst2a output from the vertical scanning circuit 77. The nodes 145 and 165 are connected to the gates of the amplifiers 147 and 167, respectively. The outputs of the amplifiers 147 and 167 are connected to the column signal line 170 via selection transistors 148 and 168, respectively. The selection transistors 148 and 168 are controlled by signals φsel1 and φsel2 output from the vertical scanning circuit 77, respectively. By sequentially switching the selection transistors 148 and 168, the pixel signals accumulated in the pixel units 50-1 and 50-2 are read out to the column processing circuit 60, respectively.

  FIG. 7 is a diagram illustrating a configuration of the column processing circuit 60 according to the first embodiment. The column processing circuit 60 includes a load 171, a programmable amplifier 172, a comparator a 173, a counter 174, a memory 175, and a comparator b 176. The load 171 is a current source, and the voltage generated in the load 171 becomes an optical signal, that is, a pixel signal. The programmable amplifier 172 amplifies the optical signal and outputs it to the comparator a173.

  The ramp waveform is input to the comparator a173. The ramp waveform is a waveform in which the horizontal axis is time t and the vertical axis is voltage V input to the comparator a 173, and the voltage V is proportional to time t. The comparator a 173 outputs a signal to the counter 174 when the input voltage, that is, the output voltage of the programmable amplifier 172 matches the ramp waveform. The counter 174 counts the time until the voltage value of the ramp waveform matches the output voltage of the programmable amplifier 172. That is, the counter value is proportional to the output voltage of the programmable amplifier 172, and the counter value increases as the output voltage of the programmable amplifier 172 increases. In other words, the counter value indicates a digital value corresponding to the output voltage of the programmable amplifier 172, that is, the optical signal.

  The counter 174 is an up / down counter that counts clock signals, and can count both in the direction of decreasing numbers and in the direction of increasing numbers. That is, the counter 174 counts in the direction in which the number is increased by the up counter, that is, in the plus direction, and counts in the direction in which the number is decreased by the down counter, that is, in the minus direction. When the counter 174 counts with the up counter, the counter value becomes a value proportional to the output voltage of the programmable amplifier 172. On the other hand, when the counter 174 counts with the down counter, the counter value is a value obtained by reversing the value of the value proportional to the output voltage of the programmable amplifier 172. The counter 174 performs a CDS (Correlated Double Sampling) process by, for example, counting a reference voltage value with a down counter and counting a voltage value of an optical signal with an up counter. In this way, the comparator a 173 and the counter 174 constitute an ADC (Analog-to-Digital Converter) circuit.

  The comparator b176 controls the operation of the counter. The comparator b 176 compares the value of the counter 174 with a predetermined threshold value and determines the next counter operation. This predetermined threshold will be described in detail later. The numerical value after the CDS processing of the counter 174 is accumulated in the memory 175. As shown in FIG. 5, signals from the plurality of column signal lines 170 are output to the horizontal scanning circuit 177, and the horizontal scanning circuit 177 sequentially receives signals indicating processing results from the plurality of column signal lines 170. To the next circuit. This expands the dynamic range. This operation will be described later. When the dynamic range is not expanded, the comparator b176 need not be used.

(Operation)
Next, the operation of the circuit of the pixel unit 50 shown in FIG. 6 will be described. As the operation mode, both a global shutter operation and a rolling shutter operation are possible. In the first embodiment, as described below, the dynamic range is expanded for both the global shutter operation and the rolling shutter operation. It is possible.

(Global shutter operation)
FIG. 8 is a diagram showing a timing chart of the global shutter operation. The global shutter mode includes a step of accumulating signals in a capacitor at all pixels and a step of sequentially reading out signals accumulated in the capacitor. The left side of the timing chart shown in FIG. 8 is an accumulation process, and the right side is a reading process. The accumulation process is performed simultaneously for all the pixel units 50. On the other hand, the reading process is sequentially performed for each pixel unit 50 by switching the pixel unit 50 electrically connected to the column signal line 170 by the selection transistors 148 and 168.

  First, at time g1, all the pixel units 50 are reset all at once. Specifically, when the signals φrst1 and φrst2 become high, the floating diffusions 133 and 153 are reset, respectively. Next, at time g2, when the signals φtx1 and φtx2 become high, the electrons accumulated in the photodiodes 130 and 150 are transferred to the floating diffusions 133 and 153, respectively. Thereby, the electrons in the photodiodes 130 and 150 become empty. Thereafter, exposure, that is, photoelectric conversion is performed so that light enters the photodiodes 130 and 150 in this state. This photoelectric conversion is performed from the end of time g2 until the next transfer of electrons from the photodiodes 130 and 150 to the floating diffusions 133 and 153 to complete the transfer, that is, until the end of time t3. The period from the end of time g2 to the end of time t3 is defined as a first period T1.

  Next, at time t1, all the pixel units 50 are reset all at once. Specifically, when the signals φrst1 and φrst2 become high, the floating diffusions 133 and 153 are reset, respectively. At this time, the capacitors 142 and 144 are reset when the signals φrst1a, φtx1b, and φtx1c become high. Similarly, at this time, the capacitors φrst2a, φtx2b, and φtx2c become high, whereby the capacitors 162 and 164 are reset. Note that signals are written to the capacitors 142, 144, 162, and 164 by voltages generated by the amplifier transistors 134 and 154 in the load transistors 135 and 155, respectively. Therefore, it is not always necessary to reset. However, when writing to each capacitor, there may be a problem in that there is a slight difference between when writing in a direction from lower voltage to higher voltage and when writing in a direction from higher voltage to lower voltage. In order to avoid such a problem, resetting is performed so that each capacitor is set in a higher voltage direction.

  The potentials in the floating diffusions 133 and 153 reset as described above become the reference potentials of the pixel units 50-1 and 50-2, respectively. This reference potential is written into the capacitors 142 and 162 at time t2. Specifically, when the signals φld1 and φld2 become high and a predetermined voltage is applied to the gates of the load transistors 135 and 155, respectively, the load transistors 135 and 155 operate as current sources. As a result, current flows through the amplifier transistors 134 and 154, whereby the amplifier transistors 134 and 154 operate as source follower circuits, respectively. Further, at this time, the signals φtx1a and φtx1b become high, whereby the reference potential is recorded in the capacitor 142 via the connection portion 136. Similarly, when the signals φtx2a and φtx2b become high, the reference potential is recorded in the capacitor 162 via the connection unit 156.

  At this time, as described above, the charges (electrons in the first embodiment) generated by the photoelectric conversion performed during the first period T1 are accumulated in the photodiodes 130 and 150. At time t3, when the signals φtx1b and φtx2b become low and the signals φtx1 and φtx2 become high, charges are transferred from the photodiodes 130 and 150 to the floating diffusions 133 and 153, respectively. As a result, the potentials of the floating diffusions 133 and 153 change by the amount of the transferred charge. For example, when the charge is an electron, the potentials of the floating diffusions 133 and 153 are lowered. This changed potential becomes the potential of the optical signal of the pixel units 50-1 and 50-2 in the first period T1.

  Next, when the signals φtx1c and φtx2c become high at time t4, the potentials of the floating diffusions 133 and 153 after the charge transfer are amplified by the amplifier transistors 134 and 154 and recorded in the capacitors 144 and 164, respectively. After the recording, the signals φld1, φtx1a, φtx1c and φld2, φtx2a, φtx2c become low, and the recording of the optical signal in the first period T1 is completed.

  At time t3, when the transfer of charges from the photodiodes 130 and 150 to the floating diffusions 133 and 153 is completed and the signals φtx1 and φtx2 become low, the next exposure, that is, photoelectric conversion, starts in the photodiodes 130 and 150. . Thereafter, at time w1, the signals φrst1 and φrst2 become high, thereby resetting the floating diffusions 133 and 153, respectively. Thereafter, at time w2, the signals φtx1 and φtx2 become high, whereby charges are transferred from the photodiodes 130 and 150 to the floating diffusions 133 and 153, respectively. Note that the period from the end of time t3 to the end of time w2 is the second period T2.

  Here, in the first embodiment, the first period T1 is configured to be longer than the second period T2. In the case where the times g1 and g2 are not provided, the first period T1 is a maximum from the end of the time w2 to the time t3 when the charge is transferred by setting the signal φtx1 high in the next frame. It will be time. Accordingly, the first optical signal obtained by photoelectric conversion in the first period T1 which is a long period is accumulated in the capacitor 144 and the capacitor 164 at time t4. Further, the reference potential of the first optical signal is stored in the capacitor 142 and the capacitor 162. On the other hand, the second optical signal obtained by photoelectric conversion in the second period T2, which is a period shorter than the first period T1, is accumulated in the floating diffusion 133 and the floating diffusion 153 at time w2. That is, the first optical signal is an optical signal obtained by long-time exposure, and the second optical signal is an optical signal obtained by short-time exposure.

  As described above, in all the pixel units 50, the optical signals obtained by performing the photoelectric conversion in the first period T1 are recorded in the capacitors 142 and 144 and the capacitors 162 and 164 provided on the substrate 2. On the other hand, the optical signal obtained by photoelectric conversion in the second period T2 is recorded in the floating diffusion 133 provided on the substrate 1 and in the floating diffusion 153. Next, these recorded optical signals are read out. Unlike the recording process, the reading process is not performed at the same time for all the pixel units 50, but sequentially for each pixel unit 50 connected to the column signal line 170.

  As shown in FIG. 7, a column processing circuit 60 having a load and an ADC circuit is connected to the column signal line 170, and the column processing circuit 60 processes a signal transmitted through the column signal line 170. The process will be described with reference to the timing chart of FIG. 8 and the flowchart of FIG.

  FIG. 9 is a flowchart illustrating an operation when the global shutter operation of the column processing circuit 60 according to the first embodiment is performed. First, at time t15, when the signal φsel1 becomes high, the selection transistor 148 is turned on and the pixel unit 50-1 is selected. At this time, the node 145 is reset by the signal φrst1a becoming high. Thereafter, at time t16, the signal φtx1b becomes high, whereby the voltage level of the capacitor 142 is recorded in the node 145. Specifically, the reference voltage is recorded in the node 145 by resetting the node 145 at time t15. Here, the capacitance of the capacitor 142 is sufficiently larger than the stray capacitance of the node 145. Therefore, when the node 145 in which the reference voltage has been recorded is charged by the capacitor 142 at time t16, the voltage level of the node 145 becomes substantially the same as the voltage level of the capacitor 142.

  At this time, the potential of the node 145 is transmitted to the column signal line 170 via the amplifier 147. In other words, the voltage level of the capacitor 142 is transmitted to the column signal line 170. The voltage generated in the load 171 is amplified by the programmable amplifier 172 and transmitted to the comparator a173. At this time, the ramp waveform is input to the comparator a173. Further, the counter 174 is in a state of being reset in advance before measurement.

  When the ramp waveform is input to the comparator a173, the counter 174 counts the reference voltage value Vstd1 obtained by amplifying the reference potential accumulated in the capacitor 142 by the programmable amplifier 172 in the down direction. Thereafter, when the value output from the programmable amplifier 172 matches the value of the ramp waveform, the comparator a 173 outputs a pulse to the counter 174, whereby the counter 174 stops the counter operation (S102). In this way, a counter value corresponding to a value obtained by reversing the sign of the reference voltage value Vstd1 is obtained.

  Next, at time t17, the signal φrst1a goes high, whereby the node 145 is reset again. Next, at time t18, the signal φtx1c becomes high, so that the voltage level of the capacitor 144 on which the optical signal is recorded is output to the column signal line 170 via the amplifier 147 in the same manner as the processing at time t16. Is done.

  Here, in the column processing circuit 60, when the ramp waveform is supplied to the comparator a173, the counter 174 is obtained by amplifying the voltage level of the first optical signal accumulated in the capacitor 144 by the programmable amplifier 172. The optical signal voltage value Vopt1 is counted in the up direction. Thereafter, when the value output from the programmable amplifier 172 matches the value of the ramp waveform, the comparator a 173 outputs a pulse to the counter 174, whereby the counter 174 stops the counter operation (S103). In this way, a counter value corresponding to the optical signal voltage value Vopt1 is obtained.

  Here, the counter 174 counts the reference voltage value Vstd1 in the down direction at time t16, that is, S102, and counts the optical signal voltage value Vopt1 in the up direction at time t17, that is, S103. Therefore, the counter 174 counts the difference between the reference voltage value Vstd1 and the optical signal voltage value Vopt1. That is, the CDS process has been performed. In this way, the counter 174 obtains a counter value corresponding to a value obtained by subtracting the reference voltage from the voltage value of the first optical signal obtained by photoelectric conversion in the first period T1.

  When the reading of the first optical signal obtained in the first period T1 is completed, the comparator b 176 compares the counter value obtained by the counter 174 with a predetermined voltage threshold Th1v. That is, the comparator b 176 determines whether or not the counter value obtained by the counter 174 is equal to or greater than the voltage threshold Th1v (S104). The comparator b 176 outputs the second optical signal obtained in the second period T2 when the value of the first optical signal obtained in the first period T1 is equal to or greater than the voltage threshold Th1v (YES in S104). The counter 174 is instructed to perform an addition operation, that is, a synthesis operation. On the other hand, when the value of the first optical signal obtained in the first period T1 is smaller than the voltage threshold Th1v (NO in S104), the comparator b176 outputs the second optical signal obtained in the second period T2. The counter 174 is instructed not to perform the addition operation, that is, the synthesis operation.

  Here, when the first optical signal obtained in the first period T1 is equal to or higher than the voltage threshold Th1v, the voltage threshold Th1v is the noise of the first optical signal obtained in the first period T1. It is determined to be equal to or higher than the noise of the second optical signal obtained in the second period T2. Hereinafter, the voltage threshold Th1v will be described with reference to FIG.

  FIG. 10 is a diagram comparing the noise of the first optical signal and the noise of the second optical signal. In FIG. 10, a thick solid line indicates the noise of the first optical signal, and a thick broken line indicates the noise of the second optical signal. In FIG. 10, the horizontal axis indicates the intensity of each optical signal, and specifically indicates the number of electrons after photoelectric conversion of each optical signal. The vertical axis represents noise in each optical signal. Specifically, the noise is converted into the number of electrons.

  The second optical signal obtained in the second period T2 is recorded in the floating diffusion 133. Here, the noise generated in the floating diffusion 133 is very large. For example, light contamination from which light incident from the back surface of the substrate 1 (upper side in FIG. 4) reaches the floating diffusion 133 occurs, and further, the signal decreases due to leakage current, but this light contamination becomes noise. Optical contamination increases as the intensity of the optical signal increases. Alternatively, since the reset potential when recording in the floating diffusion 133 is not stored, KTC noise, which is thermal noise, is generated. In addition, noise due to dark current is also generated. The total of these noises is, for example, about 50 to 100 electrons as shown by the thick broken line in FIG. Therefore, when the second optical signal including such noise is combined with the first optical signal as it is in order to expand the dynamic range, the pixel signal becomes an optical signal having a large noise as a whole.

  On the other hand, the first optical signal obtained in the first period T1 is recorded in the capacitor 144. Noise is also generated in the first optical signal. The noise generated in the first optical signal includes, for example, noise due to capacitor leakage current and shot noise. In the first embodiment, shot noise is considered as noise generated in the first optical signal. Shot noise is equal to the square root of the optical signal. That is, the shot noise increases as the intensity of the optical signal increases, as indicated by the thick solid line in FIG.

  As shown in FIG. 10, when the intensity of the optical signal is small, the noise of the first optical signal, that is, the shot noise, is smaller than the noise of the second optical signal recorded in the floating diffusion 133. At this time, if the second optical signal is combined with the first optical signal, the influence of noise of the second optical signal appears in the image quality. On the other hand, when the intensity of the optical signal is large, the noise of the first optical signal, that is, the shot noise is larger than the noise of the second optical signal recorded in the floating diffusion 133. At this time, even if the second optical signal is combined with the first optical signal, the influence of the noise of the second optical signal appearing on the image quality is small. Therefore, in order to expand the dynamic range, the second optical signal may be combined with the first optical signal when the noise of the first optical signal is greater than or equal to the noise of the second optical signal.

Here, as shown in FIG. 10, when the intensity of the optical signal is the intensity threshold Th1a, the noise of the first optical signal becomes equal to the noise of the second optical signal. For example, when the noise of the floating diffusion 133 is 50 to 100 electrons, the intensity threshold Th1a is 2500 to 10,000 electrons. In FIG. 10, the intensity threshold Th1a is 10 ⁴ or 10,000. In FIG. 10, although the noise of the floating diffusion 133 is KTC noise, dark current, and optical contamination, it is not limited to this. Therefore, the intensity threshold Th1a is determined by actually measuring the light intensity and noise using the solid-state imaging device 100.

  Therefore, when the intensity of the first optical signal is greater than or equal to the intensity threshold Th1a, the noise of the first optical signal is greater than or equal to the noise of the second optical signal. The signal may be combined with the first optical signal. On the other hand, when the intensity of the first optical signal is smaller than the intensity threshold Th1a, the noise of the first optical signal is smaller than the noise of the second optical signal. The signal is not combined with the first optical signal.

  Here, the voltage threshold Th1v is determined so as to correspond to the intensity threshold Th1a. That is, the voltage threshold Th1v is determined so that the counter value of the counter 174 becomes the voltage threshold Th1v when the intensity of the optical signal accumulated in the pixel unit 50 is the intensity threshold Th1a. The counter value of the counter 174 increases as the intensity of the optical signal accumulated in the pixel unit 50 increases. Therefore, when the counter value is equal to or greater than the voltage threshold Th1v, the intensity of the first optical signal is equal to or greater than the intensity threshold Th1a. Similarly, when the counter value is smaller than the voltage threshold Th1v, the intensity of the first optical signal is smaller than the intensity threshold Th1a.

  When the counter value is equal to or greater than the voltage threshold Th1v (YES in S104), the comparator b 176 outputs a signal indicating that addition is possible to the counter 174. As a result, the counter 174 performs an addition operation. This addition operation is performed by performing the counter operations of S105 and S106 following the counter value obtained in S103 as follows. When the signals φld1 and φtx1a become high at time w13, the second optical signal stored in the floating diffusion 133 and obtained in the second period T2 is supplied to the substrate 2 via the amplifier transistor 134 and the connection portion 136. To the node 145. At this time, since the selection transistor 148 is on, the amplifier 147 outputs the second optical signal transmitted to the node 145 to the column signal line 170. That is, the potential of the node 145, that is, the potential of the second optical signal is transmitted to the column signal line 170 via the amplifier 147.

  In other words, the voltage level of the second optical signal accumulated in the floating diffusion 133 is transmitted to the column signal line 170. As described above, the voltage generated in the load 171 is amplified by the programmable amplifier 172 and transmitted to the comparator a173. At this time, the ramp waveform is input to the comparator a173. When the ramp waveform is input to the comparator a173, the counter 174 increases the optical signal voltage value Vopt2 obtained by amplifying the voltage level of the second optical signal accumulated in the floating diffusion 133 by the programmable amplifier 172. Count by direction. When the value output from the programmable amplifier 172 matches the value of the ramp waveform, the comparator a 173 outputs a pulse to the counter 174. Thereby, the counter 174 stops the counter operation (S105). In this way, a counter value corresponding to the optical signal voltage value Vopt2 is obtained. Actually, the counter value in the counter 174 at this time is set to the counter value corresponding to the value obtained by subtracting the reference voltage from the voltage value of the first optical signal obtained in S103, and further to the second optical signal. The counter value corresponding to the voltage value of the signal is added.

  Next, when φrst1 becomes high at time w14, the floating diffusion 133 is reset. That is, at this time, a potential corresponding to the reference potential is recorded in the floating diffusion 133. Next, at time w15, the reference potential is transmitted to the node 145 of the substrate 2 through the amplifier transistor 134 and the connection portion 136. At this time, since the selection transistor 148 is on, the amplifier 147 outputs the reference potential transmitted to the node 145 to the column signal line 170 in the same manner as w13 and S105. That is, the potential of the node 145, that is, the reference potential is transmitted to the column signal line 170 via the amplifier 147.

  Here, the counter 174 of the column processing circuit 60 counts the reference voltage value Vstd2 obtained by amplifying the reference potential stored in the floating diffusion 133 by the programmable amplifier 172 in the down direction. When the value output from the programmable amplifier 172 matches the value of the ramp waveform, the comparator a 173 outputs a pulse to the counter 174. Thereby, the counter 174 stops the counter operation (S106).

  In this way, the CDS process for the second optical signal obtained in the second period T2 is completed. Further, the counter value of the counter 174 at this time is such that the first optical signal obtained in the first period T1 and the second optical signal obtained in the second period T2 are the end of the CDS process, respectively. The value corresponds to the value added in the state.

  On the other hand, when the counter value of the counter 174 when S103 ends is smaller than the voltage threshold Th1v (NO in S104), the addition operation is not performed. That is, the comparator b 176 outputs a signal indicating that addition cannot be performed to the counter 174. As a result, the counter 174 stops operating while the second optical signal is being output to the column signal line 170, that is, during times w13 to w15. Therefore, only the processing result of the first optical signal obtained in the first period T1 remains in the counter 174.

  Next, the counter 174 moves the counter value to the memory 175. The memory 175 stores a counter value in a state where the first optical signal and the second optical signal are combined (when passing through S105 and S106), or a counter value of only the first optical signal (S105 and S106). (If not routed) is stored (S107). Finally, the counter value accumulated in the memory 175 is read out by the horizontal scanning circuit 177 (S108). The horizontal scanning circuit 177 transmits a counter value corresponding to the digital value of the pixel signal to the next circuit for each column. Thereby, the process for one line of pixels is completed.

  Next, at time t25 to w25, the same process as at time t15 to w25 described above is performed for the next pixel unit 50-2. That is, a counter value corresponding to the pixel signal for the pixel unit 50-2, that is, a digital value is read. In this way, when the pixel signals are read out for all the pixel units 50 connected to the column signal line 170, the signal reading in all of one image is completed.

  FIG. 11 is a graph showing the relationship between the light amount and the optical signal in the first embodiment. In FIG. 11, the horizontal axis indicates the amount of light accumulated in the photodiode 130, and the vertical axis indicates the voltage level of the optical signal read from the pixel unit 50. A solid line indicates an optical signal read from the pixel unit 50. A broken line indicates an optical signal obtained by exposure for a long time (first period T1) or an optical signal obtained by exposure for a short time (second period T2). As shown in FIG. 11, the optical signal increases in proportion to the amount of light, but when the upper limit of the dynamic range is reached, the optical signal is saturated at the voltage level of a certain optical signal.

  On the other hand, in the first embodiment, as shown by the solid line in FIG. 11, the light to be read is read until the voltage level of the first optical signal obtained in the first period T1 reaches a predetermined threshold value. The signal is only the first optical signal obtained in the first period T1. Here, the “predetermined threshold value” is a voltage level corresponding to the voltage threshold value Th1v. Then, when the voltage level of the first optical signal obtained in the first period T1 reaches a predetermined threshold value, the second light obtained in the second period T2 is added to the first optical signal. An optical signal obtained by combining the signals is read out. Then, even after the voltage level of the first optical signal obtained in the first period T1 is saturated, the first optical signal and the second optical signal are combined as the amount of light increases. The optical signal continues to increase. As a result, the dynamic range can be expanded.

  As described above, in the first embodiment, in the case of the global shutter operation, the second optical signal obtained by the second period T2, that is, the short time exposure, is accumulated in the floating diffusions 133 and 153. Here, the floating diffusions 133 and 153 are minute capacities constituted by diffusion layers, and are also provided in existing solid-state imaging devices. Therefore, the solid-state imaging device 100 according to the first embodiment can expand the dynamic range without increasing the number of capacitors.

  Further, as described above, since the optical signal is synthesized only when the first optical signal is equal to or greater than a predetermined threshold, the influence of noise is increased by synthesizing the second optical signal. It can also be suppressed. In other words, if the floating diffusion is simply used to store the optical signal obtained by short-time exposure, the image quality may deteriorate due to the synthesis of the optical signal stored in the floating diffusion due to noise generated in the floating diffusion. There is. As a result, the floating diffusion cannot be used as an optical signal storage means. On the other hand, in the first embodiment, the optical signal accumulated in the floating diffusion is synthesized when the influence of the noise of the optical signal accumulated in the floating diffusion does not appear in the image quality. Therefore, in the first embodiment, it is possible to suppress an increase in the influence of noise even if the floating diffusion is used. That is, in the first embodiment, it is possible to expand the dynamic range while suppressing an increase in the influence of noise without increasing the number of capacitors.

(Rolling shutter operation)
Next, the rolling shutter operation will be described. In the rolling shutter operation, processing is not performed on all pixels at the same time as in the global shutter operation, but processing is performed for each row. In the global shutter operation, since exposure is performed at the same time for all the pixels, the time information of the pixel signals is aligned for all the pixels. Therefore, an image without distortion is obtained in the global shutter operation.

  On the other hand, in general, the rolling shutter operation is characterized by less noise than the global shutter operation. In the global shutter operation, exposure is performed on all pixels at once and pixel signals are accumulated all at once, and then the accumulated pixel signals are sequentially read out row by row. Therefore, in the pixel located in the lower row, the time for holding the pixel signal becomes longer. The longer the holding time is, the more noise is superimposed on the held signal charge, and the image quality is degraded. On the other hand, in the case of the rolling shutter operation, the exposure is sequentially performed for each row and the pixel signals are sequentially accumulated, and then the accumulated pixel signals are sequentially read for each row. Therefore, it is not necessary to hold the pixel signal in the storage unit, or the holding time may be short. Thereby, noise can be reduced.

  Also in the first embodiment, by performing the rolling shutter operation, the time for holding the pixel signal can be shortened compared to the case of the global shutter operation. Therefore, it is possible to reduce noise superimposed on the held pixel signal. Therefore, it is possible to selectively use a rolling shutter operation when imaging a subject whose image distortion does not matter, and a global shutter when imaging a subject whose image distortion is to be avoided.

  Hereinafter, a method for expanding the dynamic range when performing the rolling shutter operation in the first embodiment will be described. Even when the rolling shutter operation is performed, the circuit configuration of the solid-state imaging device 100 is the same as that shown in FIGS. That is, both the global shutter operation and the rolling shutter operation can be performed with the same circuit configuration.

  FIG. 12 is a timing chart of the rolling shutter operation. FIG. 13 is a flowchart illustrating an operation when the column shutter circuit 60 according to the first embodiment performs a rolling shutter operation. In the rolling shutter operation described below, processing for the pixel unit 50-1 is performed, and then processing for the pixel unit 50-2 is performed.

  First, at time g11, the signal φrst1 goes high, whereby the floating diffusion 133 is reset. Next, at time g12, the signal φtx1 becomes high, whereby the electrons accumulated in the photodiode 130 are transferred to the floating diffusion 133. Thereby, the electrons in the photodiode 130 become empty. Thereafter, exposure, that is, photoelectric conversion is performed so that light enters the photodiode 130 in this state. This photoelectric conversion is performed from the end of time g12 until the next transfer of electrons from the photodiode 130 to the floating diffusion 133 to complete the transfer, that is, until the end of time r103. The period from the end of time g12 to the end of time r103 is the first period T1.

  On the other hand, the floating diffusion 153 is reset when the signal φrst2 becomes high at time g21. Next, at time g <b> 22, the signal φtx <b> 2 becomes high, so that the electrons accumulated in the photodiode 150 are transferred to the floating diffusion 153. Thereby, the electrons in the photodiode 150 become empty. Thereafter, exposure, that is, photoelectric conversion is performed so that light is incident on the photodiode 150 in this state. This photoelectric conversion is performed from the end of time g22 until the next transfer of electrons from the photodiode 150 to the floating diffusion 153 and the end of transfer, that is, until the end of time r203. The period from the end of time g22 to the end of time r203 is the first period T1.

  Next, at time r101 to r104, processing for the pixel unit 50-1 is performed. Specifically, the floating diffusion 133 is reset by the signal φrst1 becoming high at time r101. At this time, the capacitors 142 and 144 are reset when the signals φrst1a, φtx1b, and φtx1c become high. Next, at time r102 to r104, the signals φld1 and φtx1a become high. Further, when the signal φtx1b becomes high at time r102, the reset potential of the floating diffusion 133, that is, the reference potential is amplified by the amplifier transistor 134 and recorded in the capacitor 142.

  At this time, as described above, the photodiode 130 accumulates charges generated by the photoelectric conversion performed during the first period T1. At time r103, when the signal φtx1b becomes low and the signal φtx1 becomes high, the charge (here, electrons) of the photodiode 130 is transferred to the floating diffusion 133, and the potential of the floating diffusion 133 changes. Next, at time r104, when the signal φtx1 becomes low and the signal φtx1c becomes high, the potential after the charge transfer of the floating diffusion 133 is amplified by the amplifier transistor 134 and recorded in the capacitor 144. In this way, the first optical signal obtained in the first period T1 is recorded in the capacitors 142 and 144, that is, the first optical signal storage means.

  In the photodiode 130, the second optical signal is accumulated in the second period T2 after the time r104 after the charge transfer at the time r103. The second period T2 is a period from the end of the time r103 until the next transfer of electrons from the photodiode 130 to the floating diffusion 133 to complete the transfer, that is, the end of the time r110. Here, also in the rolling shutter operation, similarly to the global shutter operation described above, the first period T1 is configured to be longer than the second period T2.

  On the other hand, in the time r201 to r204, the first optical signal obtained in the first period T1 for the pixel unit 50-2 in the same manner as the time r101 to r104 is the first optical signal storage unit (specifically, Is recorded in the capacitors 162 and 164). Also in the pixel unit 50-2, after the time r204 after the charge transfer in r203, the second optical signal is accumulated in the second period T2 by the photodiode 150.

  When the predetermined time has passed, the pixel portion 50-1 reads out a signal. When the signal φsel1 becomes high at time r105, the pixel portion 50-1 is connected to the column signal line 170. At this time, when the signals φrst1 and φrst1a become high, the floating diffusion 133 and the node 145 are reset. As a result, a potential corresponding to the reference potential is recorded in the floating diffusion 133. Next, when the signal φtxb1 becomes high at time r106, the potential of the capacitor 142 is transmitted to the column signal line 170 via the amplifier 147.

  In the column processing circuit 60 connected to the column signal line 170, in S202 of FIG. 13, as in S102, the counter 174 obtains the reference potential obtained by amplifying the reference potential stored in the capacitor 142 by the programmable amplifier 172. The voltage value Vstd1 is counted in the down direction. Next, at time r107, the signal φrst1a becomes high, whereby the node 145 is reset. Next, at time r108, the signal φtx1c becomes high, whereby the potential of the capacitor 144, that is, the potential of the first optical signal is transmitted to the column signal line 170 via the amplifier 147. At this time, in S203 of FIG. 13, as in S103, the counter 174 causes the optical signal voltage value Vopt1 obtained by amplifying the voltage level of the first optical signal stored in the capacitor 144 by the programmable amplifier 172. Is counted in the up direction. As a result, the difference between the reference voltage value Vstd1 stored in the capacitor 142 and the optical signal voltage value Vopt1 stored in the capacitor 144 is subjected to CDS processing and corresponds to the first optical signal obtained in the first period T1. Counter value to remain in the counter.

  Next, in S204 of FIG. 13, the comparator b176 determines whether or not the counter value is equal to or higher than a voltage threshold Th2v that is a predetermined threshold. What is important here is that the voltage threshold Th2v is different from the voltage threshold Th1v which is a threshold in the case of the global shutter operation. In the case of the global shutter operation, the second optical signal obtained in the second period T2 is stored in the floating diffusions 133 and 153. In this case, since the noise due to the KTC noise and the leakage current is large, the voltage threshold Th1v becomes the threshold. On the other hand, in the case of the rolling shutter operation, the second optical signal obtained in the second period T2 can be stored in the photodiodes 130 and 150. Here, noise with respect to electrons accumulated in the photodiodes 130 and 150 is much smaller than noise in the floating diffusions 133 and 153. Therefore, the voltage threshold Th2v may be smaller than the voltage threshold Th1v. The noise value is 50 to 100 in the case of the global shutter operation, whereas it is 5 to 10 in the case of the rolling shutter operation. As a result, the intensity threshold Th2A corresponding to the voltage threshold Th2v is 25 to 100 in number of electrons, which is two orders of magnitude smaller than the intensity threshold Th1A (about 10000) in the global shutter operation.

  In this way, when the counter value is compared with the voltage threshold Th2v and the first optical signal, that is, the counter value obtained in the first period T1 after the CDS process is larger (YES in S204), the counter 174 When the second optical signal obtained in the second period T2 is read, the addition operation is performed. This addition operation is performed by further performing the counter operations of S205 and S206 following the counter value obtained in S203 as follows.

  When the signals φld1 and φtx1a become high at time r109, the potential of the floating diffusion 133 that is reset and accumulated at time r105 is transmitted to the node 145 through the amplifier transistor 134. In other words, the reference potential accumulated in the floating diffusion 133 is transmitted to the node 145. At this time, since the selection transistor 148 is on, the reference potential transmitted to the node 145 is output to the column signal line 170 via the amplifier 147 as in the case of w15 and S106. The counter 174 of the column processing circuit 60 counts the reference voltage value Vstd2 obtained by amplifying the reference potential stored in the floating diffusion 133 by the programmable amplifier 172 in the down direction (S205).

  Next, when the signal φtx1 becomes high at time r110, the charge accumulated in the photodiode 130 is transferred to the floating diffusion 133. In other words, the second optical signal stored in the photodiode 130 in the second period T2 is transferred to the floating diffusion 133. As a result, the potential of the floating diffusion 133 varies. Next, when the signal φtx1 becomes low at time r111, the charge transfer ends. Then, the potential of the floating diffusion 133 is transmitted to the node 145 through the amplifier transistor 134. Further, the potential of the node 145 is transmitted to the column signal line 170 through the amplifier 147. Similarly to w13 and S105, the counter 174 of the column processing circuit 60 increases the optical signal voltage value Vopt2 obtained by amplifying the voltage level of the second optical signal stored in the photodiode 130 in the up direction. Is counted (S206). As a result, the CDS operation is performed on the second optical signal obtained in the second period T2. Further, the counter value of the counter 174 at this time is such that the first optical signal obtained in the first period T1 and the second optical signal obtained in the second period T2 are the end of the CDS process, respectively. The value corresponds to the value added in the state.

  Note that the order of counting in the up direction and the counting in the down direction in the processes of S205 and S206 is opposite to the order of counting in the up direction and counting in the down direction in the processes of S105 and S106. That is, in the case of the global shutter operation shown in FIG. 9, the optical signal voltage value Vopt2 is counted in the up direction in S105, and then the reference voltage value Vstd2 is counted in the down direction in S106. On the other hand, in the case of the rolling shutter operation shown in FIG. 13, the reference voltage value Vstd2 is counted in the down direction in S205, and then the optical signal voltage value Vopt2 is counted in the up direction in S206. This is because the storage location of the second optical signal obtained in the second period T2 differs between the case of the global shutter operation and the case of the rolling shutter operation, and the method of reading out the second optical signal differs. Because. Specifically, in the case of the global shutter operation, the second optical signal is accumulated in the floating diffusion 133, and the reference potential is accumulated in the floating diffusion 133 after the second optical signal is read out. On the other hand, in the case of the rolling shutter operation, the second optical signal is accumulated in the photodiode 130, and the reference potential is accumulated in the floating diffusion 133 before the second optical signal is read out. Here, as shown in FIG. 6, the floating diffusion 133 is provided at a position closer to the column signal line 170 than the photodiode 130. Therefore, in the case of the global shutter operation, the second optical signal is read out first, and in the case of the rolling shutter operation, the reference potential is read out first.

  On the other hand, when the counter value of the counter 174 when S203 ends is smaller than the voltage threshold Th2v (NO in S204), the addition operation is not performed. That is, the comparator b 176 outputs a signal indicating that addition cannot be performed to the counter 174. As a result, the counter 174 stops operating while the second optical signal is being output to the column signal line 170, that is, during the times r109 to r111. Therefore, only the processing result of the first optical signal obtained in the first period T1 remains in the counter 174.

  Next, the counter 174 moves the counter value to the memory 175. The memory 175 stores the counter value in a state where the first optical signal and the second optical signal are combined (when passing through S205 and S206), or the counter value of only the first optical signal (S205 and S206). Is stored) (S207). Finally, the counter value accumulated in the memory 175 is read by the horizontal scanning circuit 177, and the horizontal scanning circuit 177 transmits the counter value corresponding to the digital value of the pixel signal to the next circuit for each column. (S208).

  Next, at time r205 to r211, processing similar to that at time r105 to r111 described above is performed for the next pixel unit 50-2. That is, a counter value corresponding to the pixel signal for the pixel unit 50-2, that is, a digital value is read. In this way, when the pixel signals are read out for all the pixel units 50 connected to the column signal line 170, the signal reading in all of one image is completed.

  In this way, in the rolling shutter operation, as in the global shutter operation, it is possible to expand the dynamic range while suppressing an increase in noise. That is, in the first embodiment, in the case of the rolling shutter operation, the optical signal obtained by the short time exposure is accumulated in the photodiodes 130 and 150. Therefore, the solid-state imaging device 100 according to the first embodiment can expand the dynamic range without increasing the number of capacitors. Furthermore, since the optical signal obtained by the short-time exposure is accumulated in the photodiodes 130 and 150 instead of the floating diffusions 133 and 153, it is possible to suppress the influence of noise on the image quality. Further, as described above, the voltage threshold Th2v, which is a threshold in the case of the rolling shutter operation, for the threshold for combining the optical signal obtained by the short exposure with the optical signal obtained by the long exposure is the global shutter. It may be shorter than the voltage threshold Th1v that is a threshold in the case of operation. Therefore, in the case of the rolling shutter operation, even if the optical signal obtained by long exposure is small, the optical signal obtained by short exposure can be combined with the optical signal obtained by long exposure. It becomes possible.

(Modification)
Note that the present invention is not limited to the above-described embodiment, and can be changed as appropriate without departing from the spirit of the present invention. For example, in the flowcharts described above, one or more of each step can be omitted as appropriate. For example, the process of S107 in FIG. Moreover, in the flowchart mentioned above, the order of each step can be changed suitably. For example, the order of S102 and S103 in FIG. 9 may be reversed. These changes are the same in FIG.

  In the embodiment described above, the first optical signal obtained in the first period T1 when the first optical signal obtained in the first period T1 is greater than or equal to a predetermined threshold value. Is more than the noise of the second optical signal obtained in the second period T2. In the first embodiment, the second optical signal is combined with the first optical signal in that case. However, the present embodiment is not limited to such a configuration. The present embodiment may be configured to actually measure the noise of each optical signal and compare the noise of the first optical signal and the noise of the second optical signal. On the other hand, actually measuring the noise of an optical signal is more difficult than simply comparing the optical signal with a threshold value. Therefore, it is possible to more easily expand the dynamic range by configuring the second optical signal to be combined with the first optical signal when the first optical signal is equal to or greater than a predetermined threshold. Is possible.

  In the first embodiment, the CDS process is performed using the comparator and the counter. However, the configuration is not limited thereto. The CDS process may be performed by a clamp circuit using a capacitor, or may be performed using a differential amplifier. Also, the AD conversion is not limited to the method according to the first embodiment described above. A pipeline type AD converter may be used.

100 Solid-state imaging device 1, 2 Substrate 50 Pixel unit 130, 150 Photo diode 131, 151 Transfer transistor 132, 152, 146, 166 Reset transistor 133, 153 Floating diffusion 142, 144, 162, 164 Capacitor 148, 168 Select transistor 170 Column Signal line 60 Column processing circuit 172 Programmable amplifier 173 Comparator a
174 Counter 176 Comparator b

Claims (7)

Photoelectric conversion means for performing photoelectric conversion;
First optical signal storage means for storing optical signals obtained by photoelectric conversion by the photoelectric conversion means;
Second optical signal storage means for storing optical signals obtained by photoelectric conversion by the photoelectric conversion means;
Combining means for combining the optical signal stored in the first optical signal storage means and the optical signal stored in the second optical signal storage means;
The first optical signal storage means stores a first optical signal obtained by photoelectric conversion in the first period by the photoelectric conversion means,
The second optical signal storage means stores the second optical signal obtained by photoelectric conversion in the second period shorter than the first period by the photoelectric conversion means,
When a global shutter operation is performed, a floating diffusion is used as the second optical signal storage unit, and the first optical signal is read out before the second optical signal,
The synthesizing unit synthesizes the first optical signal and the second optical signal when noise of the first optical signal is equal to or higher than noise of the second optical signal. A first substrate;
A second substrate facing the first substrate and electrically connected to the first substrate;
The photoelectric conversion means and the second optical signal storage means are provided on the first substrate,
The solid-state imaging device according to claim 1, wherein the first optical signal storage unit is provided on the second substrate. The solid-state imaging device according to claim 2, wherein at least one of floating diffusion and the photoelectric conversion unit is used as the second optical signal storage unit. The said synthetic | combination means synthesize | combines a said 1st optical signal and a said 2nd optical signal, when the value of a said 1st optical signal is more than a predetermined threshold value. The solid-state imaging device according to item 1. The solid-state imaging device may perform the global shutter operation and the rolling shutter operation,
The solid-state imaging device according to claim 4, wherein the threshold used when the rolling shutter operation is performed is smaller than the threshold used when the global shutter operation is performed. The solid-state imaging device may perform the global shutter operation and the rolling shutter operation,
When the global shutter operation is performed, the combining unit counts the value corresponding to the second optical signal in the up direction, and then counts the value corresponding to the reference voltage in the down direction.
6. When the rolling shutter operation is performed, the synthesizing unit counts a value corresponding to the second optical signal in an up direction after counting a value corresponding to a reference voltage in a down direction. The solid-state imaging device according to any one of the above. When the global shutter operation is performed, the second optical signal is stored in a floating diffusion used as the second optical signal storage unit, and after the second optical signal is read, the reference voltage Is accumulated in the floating diffusion,
When the rolling shutter operation is performed, the second optical signal is stored in the photoelectric conversion unit used as the second optical signal storage unit, and the reference signal is read before the second optical signal is read out. The solid-state imaging device according to claim 6, wherein a voltage is accumulated in the floating diffusion.

Images